Inference of drug off-target effects on cellular signaling using interactome-based deep learning

Many diseases emerge from dysregulated cellular signaling, and drugs are often designed to target specific signaling proteins. Off-target effects are, however, common and may ultimately result in failed clinical trials. Here we develop a computer model of the cell's transcriptional response to drugs for improved understanding of their mechanisms of action. The model is based on ensembles of artificial neural networks and simultaneously infers drug-target interactions and their downstream effects on intracellular signaling. With this, it predicts transcription factors' activities, while recovering known drug-target interactions and inferring many new ones, which we validate with an independent dataset. As a case study, we analyze the effects of the drug Lestaurtinib on downstream signaling. Alongside its intended target, FLT3, the model predicts an inhibition of CDK2 that enhances the downregulation of the cell cycle-critical transcription factor FOXM1. Our approach can therefore enhance our understanding of drug signaling for therapeutic design.

Helping parents know when to seek help for an acutely ill child: Evidence based co-development of a mobile phone app using complex intervention methodology

BACKGROUND

Acute illness accounts for the majority of episodes of illness in children under five years of age and is the age group with the highest consultation rate in general practice in the UK. The number of children presenting to emergency care is also steadily increasing, having risen beyond pre-pandemic numbers. Such high, and increasing, rates of consultation have prompted concerns about parents' level of knowledge and confidence in caring for their children when they are ill, and particularly when and how to seek help appropriately.

AIM

The ASK SNIFF collaboration research programme identified parents' need for accurate and accessible information to help them know when to seek help for a sick child in 2010. This paper presents the resulting programme of research which aimed to co-develop an evidence-based safety netting intervention (mobile app) to help parents know when to seek help for an acutely ill child under the age of five years in the UK.

METHODS

Our programme used a collaborative six step process with 147 parent and 324 health professional participants over a period of six years including: scoping existing interventions, systematic review, qualitative research, video capture, content identification and development, consensus methodology, parent and expert clinical review.

RESULTS

Our programme has produced evidence-based content for an app supported by video clips. Our collaborative approach has supported every stage of our work, ensuring that the end result reflects the experiences, perspectives and expressed needs of parents and the clinicians they consult.

CONCLUSION

We have not found any other resource which has used this type of approach, which may explain why there is no published evaluation data demonstrating the impact of existing UK resources. Future mobile apps should be designed and developed with the service users for whom they are intended.

Oral Health Education for Children: Development of a Serious Game with a User-Centered Design Approach

Background: Children can learn efficiently with well-designed serious games. The use of applications to promote health has proliferated, but there is a lack of scientific studies on educational games in oral health. Materials and Methods: We developed the Brazilian version of a British and Jordanian oral health education game for children from the perspectives of Brazilian specialists and users. This descriptive study, with a qualitative and quantitative approach, comprised three phases: I-Experts' discussion of the appropriateness of the previous version of the game to Brazil; II-Development of the first Brazilian version of the game; and III-Evaluation of the first version with 15 children from 4 to 8 years of age. Results: In Phase I, the specialists agreed with the development of the Brazilian version of the game, with minor adjustments on: advice on eating; advice on oral hygiene habits, users' age group, game characters, and game purpose. Phase II: a version with a few changes in images and recommendations, written and spoken in Brazilian Portuguese. Phase III: The global average of correct answers in the game's tasks was 75.3%, ranging from 50.0% to 100%. Children reported having fun with the game, and most understood the content and its interface; their parents found the information relevant and enjoyed the gameplay with their children. Conclusions: The Oral Health Education Game offered basic information for preventing dental caries to Brazilian children aged 4-8 years old in an interactive and fun way; it could support professionals in improving oral health education.

Noninvasive monitoring technologies to identify discomfort and distressing symptoms in persons with limited communication at the end of life: a scoping review

BACKGROUND

Discomfort and distressing symptoms are common at the end of life, while people in this stage are often no longer able to express themselves. Technologies may aid clinicians in detecting and treating these symptoms to improve end-of-life care. This review provides an overview of noninvasive monitoring technologies that may be applied to persons with limited communication at the end of life to identify discomfort.

METHODS

A systematic search was performed in nine databases, and experts were consulted. Manuscripts were included if they were written in English, Dutch, German, French, Japanese or Chinese, if the monitoring technology measured discomfort or distressing symptoms, was noninvasive, could be continuously administered for 4 hours and was potentially applicable for bed-ridden people. The screening was performed by two researchers independently. Information about the technology, its clinimetrics (validity, reliability, sensitivity, specificity, responsiveness), acceptability, and feasibility were extracted.

RESULTS

Of the 3,414 identified manuscripts, 229 met the eligibility criteria. A variety of monitoring technologies were identified, including actigraphy, brain activity monitoring, electrocardiography, electrodermal activity monitoring, surface electromyography, incontinence sensors, multimodal systems, and noncontact monitoring systems. The main indicators of discomfort monitored by these technologies were sleep, level of consciousness, risk of pressure ulcers, urinary incontinence, agitation, and pain. For the end-of-life phase, brain activity monitors could be helpful and acceptable to monitor the level of consciousness during palliative sedation. However, no manuscripts have reported on the clinimetrics, feasibility, and acceptability of the other technologies for the end-of-life phase.

CONCLUSIONS

Noninvasive monitoring technologies are available to measure common symptoms at the end of life. Future research should evaluate the quality of evidence provided by existing studies and investigate the feasibility, acceptability, and usefulness of these technologies in the end-of-life setting. Guidelines for studies on healthcare technologies should be better implemented and further developed.

An integrative multi-omics approach to characterize interactions between tuberculosis and diabetes mellitus

Tuberculosis-diabetes mellitus (TB-DM) is linked to a distinct inflammatory profile, which can be assessed using multi-omics analyses. Here, a machine learning algorithm was applied to multi-platform data, including cytokines and gene expression in peripheral blood and eicosanoids in urine, in a Brazilian multi-center TB cohort. There were four clinical groups: TB-DM(n = 24), TB only(n = 28), DM(HbA1c ≥ 6.5%) only(n = 11), and a control group of close TB contacts who did not have TB or DM(n = 13). After cross-validation, baseline expression or abundance of MMP-28, LTE-4, 11-dTxB2, PGDM, FBXO6, SECTM1, and LINCO2009 differentiated the four patient groups. A distinct multi-omic-derived, dimensionally reduced, signature was associated with TB, regardless of glycemic status. SECTM1 and FBXO6 mRNA levels were positively correlated with sputum acid-fast bacilli grade in TB-DM. Values of the biomarkers decreased during the course of anti-TB therapy. Our study identified several markers associated with the pathophysiology of TB-DM that could be evaluated in future mechanistic investigations.

Use of noisy labels as weak learners to identify incompletely ascertainable outcomes: A Feasibility study with opioid-induced respiratory depression

Objective

Assigning outcome labels to large observational data sets in a timely and accurate manner, particularly when outcomes are rare or not directly ascertainable, remains a significant challenge within biomedical informatics. We examined whether noisy labels generated from subject matter experts' heuristics using heterogenous data types within a data programming paradigm could provide outcomes labels to a large, observational data set. We chose the clinical condition of opioid-induced respiratory depression for our use case because it is rare, has no administrative codes to easily identify the condition, and typically requires at least some unstructured text to ascertain its presence.

Materials and methods

Using de-identified electronic health records of 52,861 post-operative encounters, we applied a data programming paradigm (implemented in the Snorkel software) for the development of a machine learning classifier for opioid-induced respiratory depression. Our approach included subject matter experts creating 14 labeling functions that served as noisy labels for developing a probabilistic Generative model. We used probabilistic labels from the Generative model as outcome labels for training a Discriminative model on the source data. We evaluated performance of the Discriminative model with a hold-out test set of 599 independently-reviewed patient records.

Results

The final Discriminative classification model achieved an accuracy of 0.977, an F1 score of 0.417, a sensitivity of 1.0, and an AUC of 0.988 in the hold-out test set with a prevalence of 0.83% (5/599).

Discussion

All of the confirmed Cases were identified by the classifier. For rare outcomes, this finding is encouraging because it reduces the number of manual reviews needed by excluding visits/patients with low probabilities.

Conclusion

Application of a data programming paradigm with expert-informed labeling functions might have utility for phenotyping clinical phenomena that are not easily ascertainable from highly-structured data.

Distance-weighted Sinkhorn loss for Alzheimer's disease classification

Traditional loss functions such as cross-entropy loss often quantify the penalty for each mis-classified training sample without adequately considering its distance from the ground truth class distribution in the feature space. Intuitively, the larger this distance is, the higher the penalty should be. With this observation, we propose a penalty called distance-weighted Sinkhorn (DWS) loss. For each mis-classified training sample (with predicted label A and true label B), its contribution to the DWS loss positively correlates to the distance the training sample needs to travel to reach the ground truth distribution of all the A samples. We apply the DWS framework with a neural network to classify different stages of Alzheimer's disease. Our empirical results demonstrate that the DWS framework outperforms the traditional neural network loss functions and is comparable or better to traditional machine learning methods, highlighting its potential in biomedical informatics and data science.

Identification of health-related problems in youth: a mixed methods feasibility study evaluating the Youth Health Report System

BACKGROUND

Because poor health in youth risk affecting their entry in adulthood, improved methods for their early identification are needed. Health and welfare technology is widely accepted by youth populations, presenting a potential method for identifying their health problems. However, healthcare technology must be evidence-based. Specifically, feasibility studies contribute valuable information prior to more complex effects-based research. The current study assessed the process, resource, management, and scientific feasibility of the Youth Health Report System prototype, developed within a youth health clinic context in advance of an intervention study.

METHODS

This mixed-methods feasibility study was conducted in a clinical setting. The process, resource, management, and scientific feasibility of the Youth Health Report System were investigated, as recommended in the literature. Participants were youth aged 16-23 years old, attending a youth health clinic, and healthcare professionals from three clinics. The youth participants used their smart phones to respond to Youth Health Report System health questions and healthcare professionals used their computer to access the results and for registration system entries. Qualitative data were collected from interviews with healthcare professionals, which were described with thematic analysis. Youth participants' quantitative Youth Health Report System data were analyzed for descriptive statistics.

RESULTS

Feasibility analysis of qualitative data from interviews with 11 healthcare professionals resulted in three themes: We expected it could be hard; Information and routines helped but time was an issue; and The electronic case report form was valuable in the health assessment. Qualitative data were collected from the Youth Health Report System. A total of 54 youth participants completed the evaluation questionnaire, and healthcare professionals retrieved information from, and made post-appointment system entries. Quantitative results revealed few missing items and acceptable data variability. An assessment template of merged qualitative and quantitative data guided a consensus discussion among the researchers, resulting in acceptable feasibility.

CONCLUSIONS

The process-, resource-, management-, and scientific feasibility aspects were acceptable, with some modifications, strengthening the potential for a successful Youth Health Report System intervention study.

Assessing mental health from registry data: What is the best proxy?

OBJECTIVE

Medical registries frequently underestimate the prevalence of health problems compared with surveys. This study aimed to determine the registry variables that can serve as a proxy for variables studied in a mental health survey.

MATERIALS AND METHODS

Prevalences of depressive symptoms, anxiety and psychoactive medication use from the 2018 Belgian Health Interview Survey (HIS) were compared with same-year prevalences from INTEGO, a Belgian primary care registry. Participants aged 15 and above were included. We assessed correlation using Spearman's rho (SR), and agreement using the intraclass correlation coefficient (ICC). We also calculated the limits of agreement (LOAs) for each comparison. HIS questions about depressive symptoms, anxiety and psychoactive medication use were compared with the following variables from INTEGO: symptom codes, diagnosis codes, free text, antidepressant/benzodiazepine prescriptions and the combinations symptom + diagnosis codes and symptom + diagnosis codes + free text, wherever relevant.

RESULTS AND DISCUSSION

Correlation between the HIS and INTEGO was generally high, except for anxiety. Agreement ranged from fair to poor, but increased when combining certain variables, by including free text, or by increasing the prescription frequency to resemble chronic use. Agreement remained poor when comparing questions about anxiety. Prevalences from INTEGO were mostly underestimates.

CONCLUSION

The external validity of medical registries can be poor, especially compared with survey data. A considerate choice of variables and prescription chronicity is needed to accurately use a registry as a surveillance tool for mental health.

Andriole - A trailblazer in medical informatics and radiology

In a traditionally male-dominated field, the journey of Dr. Andriole represents a pioneering path in the realms of radiology and medical imaging informatics. Her career has not only reshaped the landscape of radiology but also championed diversity, equity, and inclusion in healthcare technology. Through a comprehensive exploration of Dr. Andriole's career trajectory, we navigate her transition from analog to digital radiology, her influential role in pioneering picture archiving communication systems (PACS), and her dedication to mentorship and education in the field. Dr. Andriole's journey underscores the growing influence of women in radiology and informatics, exemplified by her Gold Medal accolades from esteemed organizations. Dr. Andriole's career serves as a beacon for aspiring radiologists and informaticians, emphasizing the significance of passion, mentorship, and collaborative teamwork in advancing the fields of radiology and informatics.

Ensemble machine learning reveals key features for diabetes duration from electronic health records

Diabetes is a metabolic disorder that affects more than 420 million of people worldwide, and it is caused by the presence of a high level of sugar in blood for a long period. Diabetes can have serious long-term health consequences, such as cardiovascular diseases, strokes, chronic kidney diseases, foot ulcers, retinopathy, and others. Even if common, this disease is uneasy to spot, because it often comes with no symptoms. Especially for diabetes type 2, that happens mainly in the adults, knowing how long the diabetes has been present for a patient can have a strong impact on the treatment they can receive. This information, although pivotal, might be absent: for some patients, in fact, the year when they received the diabetes diagnosis might be well-known, but the year of the disease unset might be unknown. In this context, machine learning applied to electronic health records can be an effective tool to predict the past duration of diabetes for a patient. In this study, we applied a regression analysis based on several computational intelligence methods to a dataset of electronic health records of 73 patients with diabetes type 1 with 20 variables and another dataset of records of 400 patients of diabetes type 2 with 49 variables. Among the algorithms applied, Random Forests was able to outperform the other ones and to efficiently predict diabetes duration for both the cohorts, with the regression performances measured through the coefficient of determination R2. Afterwards, we applied the same method for feature ranking, and we detected the most relevant factors of the clinical records correlated with past diabetes duration: age, insulin intake, and body-mass index. Our study discoveries can have profound impact on clinical practice: when the information about the duration of diabetes of patient is missing, medical doctors can use our tool and focus on age, insulin intake, and body-mass index to infer this important aspect. Regarding limitations, unfortunately we were unable to find additional dataset of EHRs of patients with diabetes having the same variables of the two analyzed here, so we could not verify our findings on a validation cohort.

Experimental prognostic model integrating N6-methyladenosine-related programmed cell death genes in colorectal cancer

Colorectal cancer (CRC) intricacies, involving dysregulated cellular processes and programmed cell death (PCD), are explored in the context of N6-methyladenosine (m6A) RNA modification. Utilizing the TCGA-COADREAD/CRC cohort, 854 m6A-related PCD genes are identified, forming the basis for a robust 10-gene risk model (CDRS) established through LASSO Cox regression. qPCR experiments using CRC cell lines and fresh tissues was performed for validation. The CDRS served as an independent risk factor for CRC and showed significant associations with clinical features, molecular subtypes, and overall survival in multiple datasets. Moreover, CDRS surpasses other predictors, unveiling distinct genomic profiles, pathway activations, and associations with the tumor microenvironment. Notably, CDRS exhibits predictive potential for drug sensitivity, presenting a novel paradigm for CRC risk stratification and personalized treatment avenues.

A framework to integrate artificial intelligence training into radiology residency programs: preparing the future radiologist

OBJECTIVES

To present a framework to develop and implement a fast-track artificial intelligence (AI) curriculum into an existing radiology residency program, with the potential to prepare a new generation of AI conscious radiologists.

METHODS

The AI-curriculum framework comprises five sequential steps: (1) forming a team of AI experts, (2) assessing the residents' knowledge level and needs, (3) defining learning objectives, (4) matching these objectives with effective teaching strategies, and finally (5) implementing and evaluating the pilot. Following these steps, a multidisciplinary team of AI engineers, radiologists, and radiology residents designed a 3-day program, including didactic lectures, hands-on laboratory sessions, and group discussions with experts to enhance AI understanding. Pre- and post-curriculum surveys were conducted to assess participants' expectations and progress and were analyzed using a Wilcoxon rank-sum test.

RESULTS

There was 100% response rate to the pre- and post-curriculum survey (17 and 12 respondents, respectively). Participants' confidence in their knowledge and understanding of AI in radiology significantly increased after completing the program (pre-curriculum means 3.25 ± 1.48 (SD), post-curriculum means 6.5 ± 0.90 (SD), p-value = 0.002). A total of 75% confirmed that the course addressed topics that were applicable to their work in radiology. Lectures on the fundamentals of AI and group discussions with experts were deemed most useful.

CONCLUSION

Designing an AI curriculum for radiology residents and implementing it into a radiology residency program is feasible using the framework presented. The 3-day AI curriculum effectively increased participants' perception of knowledge and skills about AI in radiology and can serve as a starting point for further customization.

CRITICAL RELEVANCE STATEMENT

The framework provides guidance for developing and implementing an AI curriculum in radiology residency programs, educating residents on the application of AI in radiology and ultimately contributing to future high-quality, safe, and effective patient care.

KEY POINTS

• AI education is necessary to prepare a new generation of AI-conscious radiologists. • The AI curriculum increased participants' perception of AI knowledge and skills in radiology. • This five-step framework can assist integrating AI education into radiology residency programs.

Gastroenterology Fellowship and Postdoctoral Training in Omics and Statistics-Part II: How Can It Be Achieved?

Data are being generated, collected, and aggregated in massive quantities at exponentially increasing rates. This "big data," discussed in depth in the first section of this two-part series, is increasingly important to understand the nuances of the gastrointestinal tract and its complex interactions and networks involving a host of other organ systems and microbes. Creating and using these datasets correctly requires comprehensive training; however, current instruction in the integration, analysis, and interpretation of big data appears to lag far behind data acquisition. While opportunities exist for those interested in acquiring the requisite training, these appear to be underutilized, in part due to widespread ignorance of their existence. Here, to address these gaps in knowledge, we highlight existing big data learning opportunities and propose innovative approaches to attain such training. We offer suggestions at both the undergraduate and graduate medical education levels for prospective clinical and basic investigators. Lastly, we categorize training opportunities that can be selected to fit specific needs and timeframes.

Clinical use of artificial intelligence products for radiology in the Netherlands between 2020 and 2022

OBJECTIVES

To map the clinical use of CE-marked artificial intelligence (AI)-based software in radiology departments in the Netherlands (n = 69) between 2020 and 2022.

MATERIALS AND METHODS

Our AI network (one radiologist or AI representative per Dutch hospital organization) received a questionnaire each spring from 2020 to 2022 about AI product usage, financing, and obstacles to adoption. Products that were not listed on www.AIforRadiology.com by July 2022 were excluded from the analysis.

RESULTS

The number of respondents was 43 in 2020, 36 in 2021, and 33 in 2022. The number of departments using AI has been growing steadily (2020: 14, 2021: 19, 2022: 23). The diversity (2020: 7, 2021: 18, 2022: 34) and the number of total implementations (2020: 19, 2021: 38, 2022: 68) has rapidly increased. Seven implementations were discontinued in 2022. Four hospital organizations said to use an AI platform or marketplace for the deployment of AI solutions. AI is mostly used to support chest CT (17), neuro CT (17), and musculoskeletal radiograph (12) analysis. The budget for AI was reserved in 13 of the responding centers in both 2021 and 2022. The most important obstacles to the adoption of AI remained costs and IT integration. Of the respondents, 28% stated that the implemented AI products realized health improvement and 32% assumed both health improvement and cost savings.

CONCLUSION

The adoption of AI products in radiology departments in the Netherlands is showing common signs of a developing market. The major obstacles to reaching widespread adoption are a lack of financial resources and IT integration difficulties.

CLINICAL RELEVANCE STATEMENT

The clinical impact of AI starts with its adoption in daily clinical practice. Increased transparency around AI products being adopted, implementation obstacles, and impact may inspire increased collaboration and improved decision-making around the implementation and financing of AI products.

KEY POINTS

• The adoption of artificial intelligence products for radiology has steadily increased since 2020 to at least a third of the centers using AI in clinical practice in the Netherlands in 2022. • The main areas in which artificial intelligence products are used are lung nodule detection on CT, aided stroke diagnosis, and bone age prediction. • The majority of respondents experienced added value (decreased costs and/or improved outcomes) from using artificial intelligence-based software; however, major obstacles to adoption remain the costs and IT-related difficulties.

Gastroenterology Fellowship and Postdoctoral Training in Omics and Statistics-Part I: Why Is It Needed?

A multitude of federally and industry-funded efforts are underway to generate and collect human, animal, microbial, and other sources of data on an unprecedented scale; the results are commonly referred to as "big data." Often vaguely defined, big data refers to large and complex datasets consisting of myriad datatypes that can be integrated to address complex questions. Big data offers a wealth of information that can be accessed only by those who pose the right questions and have sufficient technical knowhow and analytical skills. The intersection comprised of the gut-brain axis, the intestinal microbiome and multi-ome, and several other interconnected organ systems poses particular challenges and opportunities for those engaged in gastrointestinal and liver research. Unfortunately, there is currently a shortage of clinicians, scientists, and physician-scientists with the training needed to use and analyze big data at the scale necessary for widespread implementation of precision medicine. Here, we review the importance of training in the use of big data, the perils of insufficient training, and potential solutions that exist or can be developed to address the dearth of individuals in GI and hepatology research with the necessary level of big data expertise.

Risk factors for allergy documentation in electronic health record: A retrospective study in a tertiary health center in Switzerland

BACKGROUND

Most hospitals use electronic health records (EHR) to warn health care professionals of drug hypersensitivity (DH) and other allergies. Indiscriminate recording of patient self-reported allergies may bloat the alert system, leading to unjustified avoidances and increases in health costs. The aim of our study was to analyze hypersensitivities documented in EHR of patients at Lausanne University Hospital (CHUV).

METHODS

We conducted a retrospective study on patients admitted at least 24 h to CHUV between 2011 and 2021. After ethical clearance, we obtained anonymized data. Because culprit allergen could be either manually recorded or selected through a list, data was harmonized using a reference allergy database before undergoing statistical analysis.

RESULTS

Of 192,444 patients, 16% had at least one allergy referenced. DH constituted 60% of all allergy alerts, mainly beta-lactam antibiotics (BLA) (30%), NSAID (11%) and iodinated contrast media (ICM) (7%). Median age at first hospitalization and hospitalization length were higher in the allergy group. Female to male ratio was 2:1 in the allergic group. Reactions were limited to the skin in half of patients, and consistent with anaphylaxis in 6%. In those deemed allergic to BLA, culprit drug was specified in 19%, 'allergy to penicillin' otherwise. It was impossible to distinguish DH based on history alone or resulting from specialized work-up.

CONCLUSIONS

Older age, longer hospital stays, and female sex increase the odds of in-patient allergy documentation. Regarding DH, BLA were referenced in 4% of inpatient records. Specific delabeling programs should be implemented to increase data reliability and patient safety.

Telehealth improves follow-up and monitoring of age-related macular degeneration during the COVID-19 pandemic

PURPOSE

To prevent vision loss, it is important to monitor patients with age-related macular degeneration (AMD) for the development of choroidal neovascularization. The coronavirus disease 2019 (COVID-19) pandemic caused many patients to miss or delay visits. To offset those gaps in care, providers utilized telehealth (TH) to evaluate patients for symptoms of disease progression and provide health education on the importance of continuous monitoring.

METHODS

This study evaluates the impact of TH encounters on the rate of return for recommended in-person examinations for 1103 patients with non-neovascular (dry) AMD seen in an outpatient ophthalmology clinic in 2019 and due for return evaluation after the outbreak of COVID-19 in 2020. Logistic regression analysis was used to identify demographic, clinical, and sociomedical factors associated with TH utilization and in-person return.

RESULTS

422 patients (38%) utilized TH during the study period. Patients who completed a TH encounter were more likely to return for an in-person examination as compared with those who did not receive TH (OR: 1.8, CI 95%: 1.4-2.3, P < 0.001). Completing a TH visit was associated with the detection of new wet AMD (OR: 3.3, 95% CI 1.04-10.6, P = 0.043), as well as with an earlier return for those patients who were found to have disease progression (62 ± 54 days vs. 100 ± 57 days, P = 0.049).

CONCLUSION

Completing a TH visit increased the rate at which patients with dry AMD returned for recommended in-person eye examinations. In many cases, this permitted the earlier detection of wet AMD, which is linked with achieving better outcomes.

A common spectrum underlying brain disorders across lifespan revealed by deep learning on brain networks

Brain disorders in the early and late life of humans potentially share pathological alterations in brain functions. However, the key neuroimaging evidence remains unrevealed for elucidating such commonness and the relationships among these disorders. To explore this puzzle, we build a restricted single-branch deep learning model, using multi-site functional magnetic resonance imaging data (N = 4,410, 6 sites), for classifying 5 different early- and late-life brain disorders from healthy controls (cognitively unimpaired). Our model achieves 62.6 ± 1.9% overall classification accuracy and thus supports us in detecting a set of commonly affected functional subnetworks, including default mode, executive control, visual, and limbic networks. In the deep-layer representation of data, we observe young and aging patients with disorders are continuously distributed, which is in line with the clinical concept of the "spectrum of disorders." The relationships among brain disorders from the revealed spectrum promote the understanding of disorder comorbidities and time associations in the lifespan.

National survey of user-reported usability of electronic medical record software in ophthalmology in Germany

PURPOSE

A nationwide, comparative survey of the physician-reported usability of electronic medical record (EMR) software used by ophthalmologists in Germany using the System Usability Scale (SUS) as a standardized metric.

METHODS

A cross-sectional survey of members of the German Ophthalmological Society (DOG) and professional association of ophthalmologists (BVA) was conducted in May 2022. All 7788 physician members of both societies were invited to participate in an anonymous online-survey by individualized links. User-reported usability of the participants main software used for electronic medical recordkeeping was assessed using the SUS (range 0-100).

RESULTS

A total of 881 participants with 51 different EMRs completed the entire questionnaire. Mean EMR-SUS score was 65.7 (SD ± 23.5). Significant differences in mean SUS of several EMR programs were observed with a range of 31.5 to 87.2 in programs with 10 or more responses. 31.8% of all main program SUS ratings were below 50 points. Female gender was associated with 4.02 higher SUS score (95% CI 0.46-7.59). Main program SUS was positively correlated with overall work-related satisfaction and work environment SUS but negatively correlated with the number of programs in the work environment. The SUS of the entire digital work environment including all programs used daily was closely correlated with the main EMR SUS, but not the number of programs used.

CONCLUSION

Our survey revealed a fragmented pattern of EMR use by ophthalmologists in Germany with many competing software products and widely diverging mean System Usability Scale scores. A considerable share of ophthalmologists report EMR usability below what is commonly considered acceptable.

Medical informed machine learning: A scoping review and future research directions

Combining domain knowledge (DK) and machine learning is a recent research stream to overcome multiple issues like limited explainability, lack of data, and insufficient robustness. Most approaches applying informed machine learning (IML), however, are customized to solve one specific problem. This study analyzes the status of IML in medicine by conducting a scoping literature review based on an existing taxonomy. We identified 177 papers and analyzed them regarding the used DK, the implemented machine learning model, and the motives for performing IML. We find an immense role of expert knowledge and image data in medical IML. We then provide an overview and analysis of recent approaches and supply five directions for future research. This review can help develop future medical IML approaches by easily referencing existing solutions and shaping future research directions.

A network-based study reveals multimorbidity patterns in people with type 2 diabetes

Patients with type 2 diabetes mellitus (T2DM) are at a heightened risk of living with multiple comorbidities. However, the comprehension of the multimorbidity characteristics of T2DM is still scarce. This study aims to illuminate T2DM's prevalent comorbidities and their interrelationships using network analysis. Using electronic medical records (EMRs) from 496,408 Chinese patients with T2DM, we constructed male and female global multimorbidity networks and age- and sex-specific networks. Employing diverse network metrics, we assessed the structural properties of these networks. Furthermore, we identified hub, root, and burst diseases within these networks while scrutinizing their temporal trends. Our findings uncover interconnected T2DM comorbidities manifesting as emergence in clusters or age-specific outbreaks and core diseases in each sex that necessitate timely detection and intervention. This data-driven methodology offers a comprehensive comprehension of T2DM's multimorbidity, providing hypotheses for clinical considerations in the prevention and therapeutic strategies.

Metabolic stratification of human breast tumors reveal subtypes of clinical and therapeutic relevance

Extensive metabolic heterogeneity in breast cancers has limited the deployment of metabolic therapies. To enable patient stratification, we studied the metabolic landscape in breast cancers (∼3000 patients combined) and identified three subtypes with increasing degrees of metabolic deregulation. Subtype M1 was found to be dependent on bile-acid biosynthesis, whereas M2 showed reliance on methionine pathway, and M3 engaged fatty-acid, nucleotide, and glucose metabolism. The extent of metabolic alterations correlated strongly with tumor aggressiveness and patient outcome. This pattern was reproducible in independent datasets and using in vivo tumor metabolite data. Using machine-learning, we identified robust and generalizable signatures of metabolic subtypes in tumors and cell lines. Experimental inhibition of metabolic pathways in cell lines representing metabolic subtypes revealed subtype-specific sensitivity, therapeutically relevant drugs, and promising combination therapies. Taken together, metabolic stratification of breast cancers can thus aid in predicting patient outcome and designing precision therapies.

Tuberculostearic acid incorporated predictive model contributes to the clinical diagnosis of tuberculous meningitis

The conventional confirmation tests of tuberculous meningitis (TBM) are usually low in sensitivity, leading to high TBM mortality. Hence, sensitive methods for indicating the presence of bacilli are required. Tuberculostearic acid (TBSA), a constituent from Mycobacterium tuberculosis had been evaluated as a promising marker, but fails to demonstrate consistent results for definite TBM. This study retrospectively reviewed medical records of 113 TBM suspects, constructing a TBSA-combined scoring system based on multiple factors, which show sensitivity and specificity of 0.8148 and 0.8814, respectively, and the area under the receiver operating characteristic curve of 0.9010. Multivariate analyses revealed four co-predictive factors strongly associated with TBSA: extra-neural tuberculosis, basal meningeal enhancement, CSF glucose/Serum glucose <0.595, and coinfection in CNS (Total). The subsequent machine learning-based validation showed correspondent importance to factors in the TBSA model. This study demonstrates a simple scoring system to facilitate TBM prediction, yield reliable diagnoses and allow timely treatment initiation.

National Development and Implementation of a Democratized Disruptive Behavior Reporting System in Health Care

INTRODUCTION

Workplace disruptive behavior/ violence (WDBV) is underreported in health care. This study evaluated a 7-year implementation of the Disruptive Behavior Reporting System (DBRS), the most robust consolidated WDBV reporting system developed in the United States within the Veterans Health Administration (VHA).

METHODS

After implementation of the system, implementation success was measured in real time by number of reports, types of staff entering reports, time to review the reports and time between when the incident occurred and report entry.

RESULTS

Over the seven years since implementation, there has been a significant increase in reporting within DBRS with more than 50,000 reports in fiscal year (FY) 2021 up from 0 to 2014. Types of staff reporting increased to 67 from 54. The median number of days to review events in FY19 Q2 was 4.79 days and the report latency has almost completely disappeared.

DISCUSSION

DBRS was designed to democratize reporting so staff can report WDBV anytime and anywhere playing a large role in the successful implementation. The increase in total number of reported events is an indication of the success of the system as it captures data historically lost due to underreporting.

CONCLUSION

DBRS development and implementation showcases how information systems can empower front-line personnel to voice behavioral safety concerns.

AI-based disease risk score for community-acquired pneumonia hospitalization

Community-acquired pneumonia (CAP) is an acute infection involving the parenchyma of the lungs, which is acquired outside of the hospital. Population-wide real-world data and artificial intelligence (AI) were used to develop a disease risk score for CAP hospitalization among older individuals. The source population included residents in Denmark aged 65 years or older in the period January 1, 1996, to July 30, 2018. 137344 individuals were hospitalized for pneumonia during the study period for which, 5 controls were matched leading to a study population of 620908 individuals. The disease risk had an average accuracy of 0.79 based on 5-fold cross-validation in predicting CAP hospitalization. The disease risk score can be useful in clinical practice to identify individuals at higher risk of CAP hospitalization and intervene to minimize their risk of being hospitalized for CAP.

Use of electronic patient records and encrypted email patient communication among Swiss chiropractors: a population-based cross-sectional study

BACKGROUND

The implementation of electronic health information technologies is a key target for healthcare quality improvement. Among Swiss chiropractors, reliable data on the use of electronic heath information technologies and distribution of the health workforce was lacking.

OBJECTIVES

To estimate the prevalence of electronic patient record (EPR) and encrypted email communication use among Swiss chiropractors and describe the geographic distribution of chiropractors in Switzerland.

METHODS

Population-based cross-sectional study of all active practising members of the Swiss Chiropractic Association (ChiroSuisse) between 3 December 2019 and 31 January 2020. We asked about clinician and practice characteristics, EPR use for clinical record keeping, use of encrypted email for patient communication, and information on EPR and encrypted email communication products used. Multivariable logistic regression analyses assessed the associations between clinician and practice characteristics and (1) EPR use, and (2) encrypted email use.

RESULTS

Among 286 eligible Swiss chiropractors (193 [68%] men; mean age, 51.4 [SD, 11.2] years), 217 (76%) completed the survey (140 [65%] men; mean age 50.7 [11.2] years). Among respondents, 47% (95% confidence interval [CI], 40-54%) reported using an EPR in their practice, while 60% (95% CI, 54-67%) endorsed using encrypted email technology. Chiropractors aged ≥ 60 (versus those ≤ 39) years were 74% less likely to use an EPR system (OR 0.26, 95% CI 0.08 to 0.77), while clinicians from practices with 4 or more chiropractors (versus those from solo practices) were over 5 times more likely to report EPR use (OR 5.6, 2.1 to 16.5). Findings for factors associated with encrypted email use were similar. The density of chiropractors in Switzerland was 3.3 per 100,000 inhabitants.

CONCLUSIONS

As of January 2020, 286 duly licensed chiropractors were available to provide musculoskeletal healthcare in Switzerland - just under 50% of responding Swiss chiropractors used an EPR system in clinical practice, while 60% used encrypted email technology. Better implementation of EPR and electronic health information technologies in Swiss chiropractic practice is possible and encouraged for the purpose of musculoskeletal healthcare quality improvement.

Exploration of the immunogenetic landscape of hyperprogressive disease after combined immunotherapy in cancer patients

The immune-genetic changes that occur in cancer patients experiencing hyperprogressive disease (HPD) during combined immunotherapy are unclear. In this study, HPD patients with pre- and post-HPD samples and non-HPD patients with solid tumors were molecularly characterized by genetic and tumor immune microenvironment (TiME) analyses of paired samples by whole-exome sequencing, RNA sequencing, and multiplex immunofluorescence. The genetic analysis of paired samples showed that almost all the tumor driver gene mutations were preserved between pre- and post-HPD tumors. HPD patients had higher frequencies of mutations in TP53 and CNN2, and a significantly higher mutant-allele tumor heterogeneity than non-HPD patients. Tumor IL-6 mRNA was upregulated in post-HPD samples vs. pre-HPD, accompanied by a potential immune suppressive TiME with an elevated M2/M1 ratio. Salvage treatment with irinotecan plus bevacizumab was effective in one HPD patient, who experienced prolonged survival. These genetic features and TiME characteristics might help identify the features of HPD after immunotherapy.

The effect of digitalization of nursing forms in ICUs on time and cost

OBJECTIVE

Intensive Care Units are one of the areas with the lowest digitization rate. This study aims to measure the effect of digitizing medical records kept in paper forms in ICUs on time-saving and paper consumption. In our study, care forms in ICUs were transferred to digital media. In our research, care forms in ICUs were transferred to digital media.

METHODS

The time required to fill out the nursing care forms on paper and digital media was measured, the change in paper and printer costs was determined, and the results were compared. Two volunteer nurses working in the ICU of a university hospital in Istanbul measured the time it took to fill out the forms of patients on paper. Then, a future projection was made using digital form data of 5,420 care days of 428 patients hospitalized between October 2017 and September 2018. Only anonymous data of patients hospitalized in the general ICU were used, and other untempered were not included in the study.

RESULTS

When the forms were filled in digitally by the nurses, one nurse per patient per day saved 56.82 min (3.95% per day).

DISCUSSION

Health care services are provided in hospitals in Turkey with 28,353 adult intensive care beds and an occupancy rate of 68%. Based on the occupancy rate of 68%, the number of full beds is 19,280. When 56.82 min are saved per bed from the forms filled by the nurses, 760.71 care days are dedicated. Considering the salary of 1,428.67 US dollars per nurse, the savings to be achieved are estimated to be 13,040,804.8 US dollars per year.

Creating a data dictionary for pediatric autonomic disorders

PURPOSE

Whether evaluating patients clinically, documenting care in the electronic health record, performing research, or communicating with administrative agencies, the use of a common set of terms and definitions is vital to ensure appropriate use of language. At a 2017 meeting of the Pediatric Section of the American Autonomic Society, it was determined that an autonomic data dictionary comprising aspects of evaluation and management of pediatric patients with autonomic disorders would be an important resource for multiple stakeholders.

METHODS

Our group created the list of terms for the dictionary. Definitions were prioritized to be obtained from established sources with which to harmonize. Some definitions needed mild modification from original sources. The next tier of sources included published consensus statements, followed by Internet sources. In the absence of appropriate sources, we created a definition.

RESULTS

A total of 589 terms were listed and defined in the dictionary. Terms were organized by Signs/Symptoms, Triggers, Co-morbid Disorders, Family History, Medications, Medical Devices, Physical Examination Findings, Testing, and Diagnoses.

CONCLUSION

Creation of this data dictionary becomes the foundation of future clinical care and investigative research in pediatric autonomic disorders, and can be used as a building block for a subsequent adult autonomic data dictionary.

Development and validation of a collagen signature to predict the prognosis of patients with stage II/III colorectal cancer

The tumor, nodes and metastasis (TNM) classification system provides useful but incomplete prognostic information and lacks the assessment of the tumor microenvironment (TME). Collagen, the main component of the TME extracellular matrix, plays a nonnegligible role in tumor invasion and metastasis. In this cohort study, we aimed to develop and validate a TME collagen signature (CS_TME) for prognostic prediction of stage II/III colorectal cancer (CRC) and to compare the prognostic values of "TNM stage + CS_TME" with that of TNM stage alone. Results indicated that the CS_TME was an independent prognostic risk factor for stage II/III CRC (hazard ratio: 2.939, 95% CI: 2.180-3.962, p < 0.0001), and the integration of the TNM stage and CS_TME had a better prognostic value than that of the TNM stage alone (AUC_(TNM+CSTME) = 0.772, AUC _TNM = 0.687, p < 0.0001). This study provided an application of "seed and soil" strategy for prognosis prediction and individualized therapy.

FAIRness through automation: development of an automated medical data integration infrastructure for FAIR health data in a maximum care university hospital

BACKGROUND

Secondary use of routine medical data is key to large-scale clinical and health services research. In a maximum care hospital, the volume of data generated exceeds the limits of big data on a daily basis. This so-called "real world data" are essential to complement knowledge and results from clinical trials. Furthermore, big data may help in establishing precision medicine. However, manual data extraction and annotation workflows to transfer routine data into research data would be complex and inefficient. Generally, best practices for managing research data focus on data output rather than the entire data journey from primary sources to analysis. To eventually make routinely collected data usable and available for research, many hurdles have to be overcome. In this work, we present the implementation of an automated framework for timely processing of clinical care data including free texts and genetic data (non-structured data) and centralized storage as Findable, Accessible, Interoperable, Reusable (FAIR) research data in a maximum care university hospital.

METHODS

We identify data processing workflows necessary to operate a medical research data service unit in a maximum care hospital. We decompose structurally equal tasks into elementary sub-processes and propose a framework for general data processing. We base our processes on open-source software-components and, where necessary, custom-built generic tools.

RESULTS

We demonstrate the application of our proposed framework in practice by describing its use in our Medical Data Integration Center (MeDIC). Our microservices-based and fully open-source data processing automation framework incorporates a complete recording of data management and manipulation activities. The prototype implementation also includes a metadata schema for data provenance and a process validation concept. All requirements of a MeDIC are orchestrated within the proposed framework: Data input from many heterogeneous sources, pseudonymization and harmonization, integration in a data warehouse and finally possibilities for extraction or aggregation of data for research purposes according to data protection requirements.

CONCLUSION

Though the framework is not a panacea for bringing routine-based research data into compliance with FAIR principles, it provides a much-needed possibility to process data in a fully automated, traceable, and reproducible manner.

Changes in medication safety indicators in England throughout the covid-19 pandemic using OpenSAFELY: population based, retrospective cohort study of 57 million patients using federated analytics

Objective

To implement complex, PINCER (pharmacist led information technology intervention) prescribing indicators, on a national scale with general practice data to describe the impact of the covid-19 pandemic on safe prescribing.

Design

Population based, retrospective cohort study using federated analytics.

Setting

Electronic general practice health record data from 56.8 million NHS patients by use of the OpenSAFELY platform, with the approval of the National Health Service (NHS) England.

Participants

NHS patients (aged 18-120 years) who were alive and registered at a general practice that used TPP or EMIS computer systems and were recorded as at risk of at least one potentially hazardous PINCER indicator.

Main outcome measure

Between 1 September 2019 and 1 September 2021, monthly trends and between practice variation for compliance with 13 PINCER indicators, as calculated on the first of every month, were reported. Prescriptions that do not adhere to these indicators are potentially hazardous and can cause gastrointestinal bleeds; are cautioned against in specific conditions (specifically heart failure, asthma, and chronic renal failure); or require blood test monitoring. The percentage for each indicator is formed of a numerator of patients deemed to be at risk of a potentially hazardous prescribing event and the denominator is of patients for which assessment of the indicator is clinically meaningful. Higher indicator percentages represent potentially poorer performance on medication safety.

Results

The PINCER indicators were successfully implemented across general practice data for 56.8 million patient records from 6367 practices in OpenSAFELY. Hazardous prescribing remained largely unchanged during the covid-19 pandemic, with no evidence of increases in indicators of harm as captured by the PINCER indicators. The percentage of patients at risk of potentially hazardous prescribing, as defined by each PINCER indicator, at mean quarter 1 (Q1) 2020 (representing before the pandemic) ranged from 1.11% (age ≥65 years and non-steroidal anti-inflammatory drugs) to 36.20% (amiodarone and no thyroid function test), while Q1 2021 (representing after the pandemic) percentages ranged from 0.75% (age ≥65 years and non-steroidal anti-inflammatory drugs) to 39.23% (amiodarone and no thyroid function test). Transient delays occurred in blood test monitoring for some medications, particularly angiotensin-converting enzyme inhibitors (where blood monitoring worsened from a mean of 5.16% in Q1 2020 to 12.14% in Q1 2021, and began to recover in June 2021). All indicators substantially recovered by September 2021. We identified 1 813 058 patients (3.1%) at risk of at least one potentially hazardous prescribing event.

Conclusion

NHS data from general practices can be analysed at national scale to generate insights into service delivery. Potentially hazardous prescribing was largely unaffected by the covid-19 pandemic in primary care health records in England.

Development and usability evaluation of an electronic health report form to assess health in young people: a mixed-methods approach

BACKGROUND

Electronic Patient-Reported Outcomes (ePROs) have potential to improve health outcomes and healthcare. The development of health-technology applications, such as ePROs, should include the potential users and be theoretically grounded. Swedish Youth Health Clinics (YHCs) offer primarily sexual and psychological healthcare for young people aged 12 to 25 years old. Young people in healthcare settings are considered a vulnerable group. The development of a collection of Patient-Reported Outcomes (PROs) in an Electronic Health Report Form (eHRF) for identifying health and health-related problems in young people, was preceded by a qualitative interview study, exploring young people's views on using an eHRF at YHCs and which questions about health an eHRF should contain. The aim of the current study was to develop and evaluate the usability of an eHRF prototype for identifying health and health-related problems in young people visiting YHCs.

METHODS

This study used a participatory design. During the development, an expert panel consisting of eight researchers and one Information Technology worker, participated. A wide literature search was performed to find PROs to construct an eHRF prototype to cover health areas. A mixed methods usability evaluation included 14 participants (young people, healthcare professionals, and an expert panel).

RESULTS

The development resulted in an eHRF prototype, containing ten reliable and valid health questionnaires addressing mental-, physical-, and sexual health and social support, a self-efficacy question, and background questions, in total 74 items. The interviews in the usability evaluation resulted in three categories describing the usability of the eHRF: 'Captures the overall health of young people but needs clarification', 'Fun, easy, and optional and will keep young people's interest', and 'Potential contribution to improve the health consultation'. The quantitative results support the usability of the eHRF for YHCs.

CONCLUSIONS

The participatory approach contributed to development of the eHRF prototype to cover health areas adapted for the target population. The usability evaluation showed that the eHRF was usable and had the potential for self-reflection and contributions to cooperation between young people and healthcare professionals during the health consultation.

The implementation of a real time early warning system using machine learning in an Australian hospital to improve patient outcomes

BACKGROUND

Early Warning Scores (EWS) monitor inpatient deterioration predominantly using vital signs. We evaluated inpatient outcomes after implementing an Artificial Intelligence (AI) based intervention in our local EWS.

METHODS

A prior study calculated a Deterioration Index (DI) with logistic regression utilising demographics, vital signs, and laboratory results at multiple time points to predict any major adverse event (MAE-all cause mortality, ICU admission, or medical emergency team activation). The current study is a single hospital, pre-post study in Australia comparing the DI plus the existing EWS (Between the Flags-BTF) to only BTF. Data were collected on all eligible inpatients (≥ 16 years, admitted ≥ 24 hours, in general non-palliative wards). Controls were inpatients in the same hospital between January and December 2019. The DI was integrated into the electronic medical record and alerts were sent to senior ward nurse phones (July 2020 -April 2021).

RESULTS

We enrolled 28,639 patients (median age 73 years, IQR:60-83) with 52.3% female. The intervention and control groups did not show any statistically significant differences apart from reduced admissions via the emergency department in the intervention group (40.4% vs 41.6%, P=0.03). Risk for an MAE was lower in intervention than control (RR: 0.81; 95%CI: 0.74-0.89). Length of hospital stay was significantly reduced in the intervention group (3.74 days, IQR 1.84-7.26) compared to the control group (3.86 days, IQR 1.86-7.86, P=0.002) CONCLUSIONS: Implementing the DI in one hospital in Australia was associated with some improved patient outcomes. Future RCTs are needed for further validation.

Editorial: The health informatics series: Evolving with a new discipline

A brief history of the book series launched by Springer-Verlag in 1988 as Computers in Healthcare stands as a case study of its role in the development of informatics in medicine. Renamed Health Informatics in 1998, the series grew to include 121 titles as of September 2022, covering topics from dental informatics to ethics, from human factors to mobile health. An analysis of three titles now in their fifth editions reveals the evolution of content in the core disciplines of nursing informatics and health information management. Shifts in topics in the second editions of two landmark titles chart the history of the field and provide a map to the development of the computer-based health record. Metrics on the publisher's website document the reach of the series, available as e-books or chapters. The growth of the series mirrors the evolution of health informatics as a discipline, and the contributions of authors and editors from around the world are evidence of international scope.

Effects and characteristics of clinical decision support systems on the outcomes of patients with kidney disease: a systematic review

OBJECTIVES

This systematic review was conducted to investigate the characteristics and effects of clinical decision support systems (CDSSs) on clinical and process-of-care outcomes of patients with kidney disease.

METHODS

A comprehensive systematic search was conducted in electronic databases to identify relevant studies published until November 2020. Randomized clinical trials evaluating the effects of using electronic CDSS on at least one clinical or process-of-care outcome in patients with kidney disease were included in this study. The characteristics of the included studies, features of CDSSs, and effects of the interventions on the outcomes were extracted. Studies were appraised for quality using the Cochrane risk-of-bias assessment tool.

RESULTS

Out of 8722 retrieved records, 11 eligible studies measured 32 outcomes, including 10 clinical outcomes and 22 process-of-care outcomes. The effects of CDSSs on 45.5% of the process-of-care outcomes were statistically significant, and all the clinical outcomes were not statistically significant. Medication-related process-of-care outcomes were the most frequently measured (54.5%), and CDSSs had the most effective and positive effect on medication appropriateness (18.2%). The characteristics of CDSSs investigated in the included studies comprised automatic data entry, real-time feedback, providing recommendations, and CDSS integration with the Computerized Provider Order Entry system.

CONCLUSION

Although CDSS may potentially be able to improve processes of care for patients with kidney disease, particularly with regard to medication appropriateness, no evidence was found that CDSS affects clinical outcomes in these patients. Further research is thus required to determine the effects of CDSSs on clinical outcomes in patients with kidney diseases.

Explainable discovery of disease biomarkers: The case of ovarian cancer to illustrate the best practice in machine learning and Shapley analysis

OBJECTIVE

Ovarian cancer is a significant health issue with lasting impacts on the community. Despite recent advances in surgical, chemotherapeutic and radiotherapeutic interventions, they have had only marginal impacts due to an inability to identify biomarkers at an early stage. Biomarker discovery is challenging, yet essential for improving drug discovery and clinical care. Machine learning (ML) techniques are invaluable for recognising complex patterns in biomarkers compared to conventional methods, yet they can lack physical insights into diagnosis. eXplainable Artificial Intelligence (XAI) is capable of providing deeper insights into the decision-making of complex ML algorithms increasing their applicability. We aim to introduce best practice for combining ML and XAI techniques for biomarker validation tasks.

METHODS

We focused on classification tasks and a game theoretic approach based on Shapley values to build and evaluate models and visualise results. We described the workflow and apply the pipeline in a case study using the CDAS PLCO Ovarian Biomarkers dataset to demonstrate the potential for accuracy and utility.

RESULTS

The case study results demonstrate the efficacy of the ML pipeline, its consistency, and advantages compared to conventional statistical approaches.

CONCLUSION

The resulting guidelines provide a general framework for practical application of XAI in medical research that can inform clinicians and validate and explain cancer biomarkers.

Development of an integrated and comprehensive clinical trial process management system

BACKGROUND

The process of initiating and completing clinical drug trials in hospital settings is highly complex, with numerous institutional, technical, and record-keeping barriers. In this study, we independently developed an integrated clinical trial management system (CTMS) designed to comprehensively optimize the process management of clinical trials. The CTMS includes system development methods, efficient integration with external business systems, terminology, and standardization protocols, as well as data security and privacy protection.

METHODS

The development process proceeded through four stages, including demand analysis and problem collection, system design, system development and testing, system trial operation, and training the whole hospital to operate the system. The integrated CTMS comprises three modules: project approval and review management, clinical trial operations management, and background management modules. These are divided into seven subsystems and 59 internal processes, realizing all the functions necessary to comprehensively perform the process management of clinical trials. Efficient data integration is realized through extract-transform-load, message queue, and remote procedure call services with external systems such as the hospital information system (HIS), laboratory information system (LIS), electronic medical record (EMR), and clinical data repository (CDR). Data security is ensured by adopting corresponding policies for data storage and data access. Privacy protection complies with laws and regulations and de-identifies sensitive patient information.

RESULTS

The integrated CTMS was successfully developed in September 2015 and updated to version 4.2.5 in March 2021. During this period, 1388 study projects were accepted, 43,051 electronic data stored, and 12,144 subjects recruited in the First Affiliated Hospital, Zhejiang University School of Medicine.

CONCLUSION

The developed integrated CTMS realizes the data management of the entire clinical trials process, providing basic conditions for the efficient, high-quality, and standardized operation of clinical trials.

MSIMEP: Predicting microsatellite instability from microarray DNA methylation tumor profiles

Deficiency in DNA MMR activity results in tumors with a hypermutator phenotype, termed microsatellite instability (MSI). Beyond its utility in Lynch syndrome screening algorithms, today MSI has gained importance as predictive biomarker for various anti-PD-1 therapies across many different tumor types. Over the past years, many computational methods have emerged to infer MSI using either DNA- or RNA-based approaches. Considering this together with the fact that MSI-high tumors frequently exhibit a hypermethylated phenotype, herein we developed and validated MSIMEP, a computational tool for predicting MSI status from microarray DNA methylation tumor profiles of colorectal cancer samples. We demonstrated that MSIMEP optimized and reduced models have high performance in predicting MSI in different colorectal cancer cohorts. Moreover, we tested its consistency in other tumor types with high prevalence of MSI such as gastric and endometrial cancers. Finally, we demonstrated better performance of both MSIMEP models vis-à-vis a MLH1 promoter methylation-based one in colorectal cancer.

Artificial intelligence applied in pulmonary hypertension: a bibliometric analysis

Introduction

Advances in Artificial Intelligence (AI) offer new Information Technology (IT) opportunities in various applications and fields (industry, health, etc.). The medical informatics scientific community expends tremendous effort on the management of diseases affecting vital organs making it a complex disease (lungs, heart, brain, kidneys, pancreas, and liver). Scientific research becomes more complex when several organs are simultaneously affected, as is the case with Pulmonary Hypertension (PH), which affects both the lungs and the heart. Therefore, early detection and diagnosis of PH are essential to monitor the disease's progression and prevent associated mortality.

Method

The issue addressed relates to knowledge of recent developments in AI approaches applied to PH. The aim is to provide a systematic review through a quantitative analysis of the scientific production concerning PH and the analysis of the networks of this production. This bibliometric approach is based on various statistical, data mining, and data visualization methods to assess research performance using scientific publications and various indicators (e.g., direct indicators of scientific production and scientific impact).

Results

The main sources used to obtain citation data are the Web of Science Core Collection and Google Scholar. The results indicate a diversity of journals (e.g., IEEE Access, Computers in Biology and Medicine, Biology Signal Processing and Control, Frontiers in Cardiovascular Medicine, Sensors) at the top of publications. The most relevant affiliations are universities from United States of America (Boston Univ, Harvard Med Sch, Univ Oxford, Stanford Univ) and United Kingdom (Imperial Coll London). The most cited keywords are "Classification", "Diagnosis", "Disease", "Prediction", and "Risk".

Conclusion

This bibliometric study is a crucial part of the review of the scientific literature on PH. It can be viewed as a guideline or tool that helps researchers and practitioners to understand the main scientific issues and challenges of AI modeling applied to PH. On the one hand, it makes it possible to increase the visibility of the progress made or the limits observed. Consequently, it promotes their wide dissemination. Furthermore, it offers valuable assistance in understanding the evolution of scientific AI activities applied to managing the diagnosis, treatment, and prognosis of PH. Finally, ethical considerations are described in each activity of data collection, treatment, and exploitation to preserve patients' legitimate rights.

[Artificial intelligence: An introduction for clinicians]

Artificial intelligence (AI) is a growing field that has the potential to transform many areas of society, including healthcare. For a physician, it is important to understand the basics of AI and its potential applications in medicine. AI refers to the development of computer systems capable of performing tasks that typically require human intelligence, such as pattern recognition, learning from data, and decision-making. This technology can be used to analyze large amounts of patient data and to identify trends and patterns that can be difficult for human physicians to detect. This can help doctors to manage their workload more efficiently and provide better care for their patients. All in all, AI has the potential to dramatically improve the practice of medicine and improve patient outcomes. In this work, the definition and the key principles of AI are outlined, with particular focus on the field of machine learning, which has been undergoing considerable development in medicine, providing clinicians with in-depth understanding of the principles underlying the new technologies ensuring improved health care.

Quality of information regarding abnormal uterine bleeding available online

INTRODUCTION

The Internet and electronic devices with Internet access allow for a greater fluidity of information and speed of communication, especially in the field of health. Abnormal uterine bleeding (AUB) affects approximately 3-30% of women and can negatively impact their health and quality of life. Information regarding AUB that is available on the Internet may not be clear or accurate, rendering it difficult to understand and likely to result in delayed medical evaluation, which subsequently leads to worsening of the AUB.

OBJECTIVE

To evaluate the quality of the information regarding AUB currently available on the Internet, including information regarding treatments.

METHODS

The Google Trends website was searched for the most widely used English terms related to AUB. The identified descriptors were searched individually on the Google, Yahoo!, and Bing search engines. The first 10 results of each search were pre-selected and evaluated for inclusion in this study. Selected websites were categorically divided into two groups (news/magazine and academic) and individually analyzed by three experts using the DISCERN quality criteria (reliability, general quality, and quality of information) and the presence or absence of the Health on the Net Foundation Code of Conduct (HONcode®) seal.

RESULTS

Of the 168 websites included in this study, 60.1% were allocated to the news/magazine group and 39.9% were allocated to the academic group. Over half of the websites (54.2%) did not have the HONcode® quality seal. Websites in the academic group were more likely to include accurate information regarding AUB with greater reliability than websites in the news/magazine group. There were no statistical differences regarding the general quality of the websites. Most websites were rated as either moderate quality (70.8%) or low quality (28.6%). The HONcode® criterion was found to be a confounding factor of the analyses, as the grouping and quality results of websites without this seal were significantly associated. In addition, websites in the news/magazines group were 6.7 times more likely to provide low quality information than websites in the academic group (odds ratio: 6.7; 95% confidence interval: 2.1-21.4).

CONCLUSION

The information regarding AUB that is available on the Internet is of low to moderate quality. Academic websites present more reliable information of greater quality. The presence of the HONcode® seal is considered important to determine the quality of the content of a website, especially for news/magazine websites, and may help Internet users identify websites that contain more reliable information. Algorithms and applications that categorize the quality of information and the reliability of health content may be useful tools that can help patients clarify their symptoms for several conditions including AUB.

Dr. Saba's innovative contributions to nursing informatics

Dr. Virginia Saba, Ed.D., RN was known as a pioneer in nursing and nursing informatics. Yet, Dr. Saba was also a profound innovator seeking to drive change through her passion and knowledge about the science of informatics. Dr. Saba saw that her expertise needed to be shared with the nursing profession and healthcare industry. As a result, she was often one of the first to create new educational opportunities and pathways for nurses to become more educated, informed, and empowered for the betterment of patient care and visibility of the nursing profession. Her innovative leadership and solution development helped to collectively create an infrastructure of formal education, training, research, and entrepreneurial solutions necessary in today's digital environment. The nursing profession is fortunate to have had such a dedicated individual who sought to bring forward positive change in ways that have sustained and grown over the last several decades of time.

Recommendations of the International Medical Informatics Association (IMIA) on Education in Biomedical and Health Informatics: Second Revision

BACKGROUND

The purpose of educational recommendations is to assist in establishing courses and programs in a discipline, to further develop existing educational activities in the various nations, and to support international initiatives for collaboration and sharing of courseware. The International Medical Informatics Association (IMIA) has published two versions of its international recommendations in biomedical and health informatics (BMHI) education, initially in 2000 and revised in 2010. Given the recent changes to the science, technology, the needs of the healthcare systems, and the workforce of BMHI, a revision of the recommendations is necessary.

OBJECTIVE

The aim of these updated recommendations is to support educators in developing BMHI curricula at different education levels, to identify essential skills and competencies for certification of healthcare professionals and those working in the field of BMHI, to provide a tool for evaluators of academic BMHI programs to compare and accredit the quality of delivered programs, and to motivate universities, organizations, and health authorities to recognize the need for establishing and further developing BMHI educational programs.

METHOD

An IMIA taskforce, established in 2017, updated the recommendations. The taskforce included representatives from all IMIA regions, with several having been involved in the development of the previous version. Workshops were held at different IMIA conferences, and an international Delphi study was performed to collect expert input on new and revised competencies.

RESULTS

Recommendations are provided for courses/course tracks in BMHI as part of educational programs in biomedical and health sciences, health information management, and informatics/computer science, as well as for dedicated programs in BMHI (leading to bachelor's, master's, or doctoral degree). The educational needs are described for the roles of BMHI user, BMHI generalist, and BMHI specialist across six domain areas - BMHI core principles; health sciences and services; computer, data and information sciences; social and behavioral sciences; management science; and BMHI specialization. Furthermore, recommendations are provided for dedicated educational programs in BMHI at the level of bachelor's, master's, and doctoral degrees. These are the mainstream academic programs in BMHI. In addition, recommendations for continuing education, certification, and accreditation procedures are provided.

CONCLUSION

The IMIA recommendations reflect societal changes related to globalization, digitalization, and digital transformation in general and in healthcare specifically, and center on educational needs for the healthcare workforce, computer scientists, and decision makers to acquire BMHI knowledge and skills at various levels. To support education in BMHI, IMIA offers accreditation of quality BMHI education programs. It supports information exchange on programs and courses in BMHI through its Working Group on Health and Medical Informatics Education.

Predicting no-show appointments in a pediatric hospital in Chile using machine learning

The Chilean public health system serves 74% of the country's population, and 19% of medical appointments are missed on average because of no-shows. The national goal is 15%, which coincides with the average no-show rate reported in the private healthcare system. Our case study, Doctor Luis Calvo Mackenna Hospital, is a public high-complexity pediatric hospital and teaching center in Santiago, Chile. Historically, it has had high no-show rates, up to 29% in certain medical specialties. Using machine learning algorithms to predict no-shows of pediatric patients in terms of demographic, social, and historical variables. To propose and evaluate metrics to assess these models, accounting for the cost-effective impact of possible intervention strategies to reduce no-shows. We analyze the relationship between a no-show and demographic, social, and historical variables, between 2015 and 2018, through the following traditional machine learning algorithms: Random Forest, Logistic Regression, Support Vector Machines, AdaBoost and algorithms to alleviate the problem of class imbalance, such as RUS Boost, Balanced Random Forest, Balanced Bagging and Easy Ensemble. These class imbalances arise from the relatively low number of no-shows to the total number of appointments. Instead of the default thresholds used by each method, we computed alternative ones via the minimization of a weighted average of type I and II errors based on cost-effectiveness criteria. 20.4% of the 395,963 appointments considered presented no-shows, with ophthalmology showing the highest rate among specialties at 29.1%. Patients in the most deprived socioeconomic group according to their insurance type and commune of residence and those in their second infancy had the highest no-show rate. The history of non-attendance is strongly related to future no-shows. An 8-week experimental design measured a decrease in no-shows of 10.3 percentage points when using our reminder strategy compared to a control group. Among the variables analyzed, those related to patients' historical behavior, the reservation delay from the creation of the appointment, and variables that can be associated with the most disadvantaged socioeconomic group, are the most relevant to predict a no-show. Moreover, the introduction of new cost-effective metrics significantly impacts the validity of our prediction models. Using a prototype to call patients with the highest risk of no-shows resulted in a noticeable decrease in the overall no-show rate.

Exploring types of conversational agents for resolving cancer patients' questions and concerns: Analysis of 100 telephone consultations on breast cancer

OBJECTIVE

This study was conducted to investigate the types of conversational agents (CA) that can help address questions and concerns ("lay topics" [LTs]).

METHODS

We analyzed audio recordings of telephone consultations with 100 breast cancer patients and their families. (1) We identified the content and mode of expression of LTs about breast cancer raised during actual telephone consultations. (2) We checked for the presence of clue information (CI) that can help patients resolve their LTs.

RESULTS

None of the 805 LTs of the 100 callers were the same. Treatment-related questions occurred in 70 of the 100 consultations. CIs were present in 52.5% of the LTs.

CONCLUSION

The results suggest that chatbots (a type of CA) that offer CIs are more feasible than chatbots that answer each question directly in cancer consultations. Moreover, it is difficult to answer questions directly because preparing answers to all LTs in a breast cancer consultation is challenging owing to LT differences. Therefore, preparing high-quality CIs focused on treatments is required.

PRACTICE IMPLICATIONS

An increasing number of cancer patients are seeking information to resolve their LTs. CAs can help supplement the limited human resources available if they are supplied with appropriate CIs.

Use of an open-source electronic health record to establish a "virtual hospital": A tale of two curricula

BACKGROUND

The electronic health record (EHR) is central to medical informatics. Its use is also recognized as an important skill for future clinicians. Typically, medical students' first exposure to an EHR is when they start their clinical internships, and medical informatics students may or may not get experience with an EHR before graduation. We describe the process of implementing an open-source EHR in two curricula: Medicine and Medical informatics. For medical students, the primary goals were to allow students to practice analyzing information from the EHR, creating therapeutic plans, and communicating with their colleagues via the EHR before they start their first clinical rotations. For medical informatics students, the primary goal was to give students hands-on experience with creating decision support in an EHR.

APPROACH

We used the OpenMRS electronic health record with a custom decision support module based on Arden Syntax. Medical students needed a secure, stable environment to practice medical reasoning. Medical informatics students needed a more isolated system to experiment with the EHR's internal configuration. Both student groups needed synthetic patient cases that were realistic, but in different aspects. For medical students, it is essential that these cases are clinically consistent, and events unfold in a logical order. By contrast, synthetic data for medical informatics students should mimic the data quality problems found in real patient data.

OUTCOMES

Medical informatics students show more mature reasoning about data quality issues and workflow integration than prior to using the EHR. Comments on both course evaluations have been positive, including comments on how working with a real-world EHR provides a realistic experience.

CONCLUSION

The open-source EHR OpenMRS has proven to be a valuable addition to both the medicine and medical informatics curriculum. Both sets of students experience use of the EHR as giving them valuable, realistic learning experiences.

[Health enabling technologies and intelligent environments in rehabilitation]

BACKGROUND

The increasing digitalization of society is having an impact on medicine. People increasingly use digital devices and services for various purposes (e.g., sports, security, convenience). Ubiquity, a strong degree of connectivity and high context sensitivity are creating intelligent environments that generate data about individuals. Suitable evaluation algorithms can extract information about the personal health status that can be used for diagnostics and treatment. Gamification methods allow patients to be more actively involved in their recovery, which can have a positive effect on adherence. Particularly in the field of rehabilitation medicine, which often affects and interacts with the personal living environment, the use of this information can make a difference.

OBJECTIVE

Using specific examples of the application of assistive health technologies and intelligent environments in rehabilitation medicine, the current state of development is presented and the possible future research directions and needs for action in this field are presented in a practical way.

MATERIAL AND METHODS

Three exemplary research projects introduce the topic, are embedded in the current state of research and allow a projection into the future against the background of many years of experience.

RESULTS

The reported projects show not only the technical feasibility but also individually the medical effectiveness of interventions.

CONCLUSION

Finally, an analysis of the barriers that have so far prevented a more intensive use of the technologies and how these might be countered is carried out.

eHealth competence building for future doctors and nurses - Attitudes and capabilities

BACKGROUND

Digitalisation is rapidly changing health care processes and the health care sector, thus increasing the need to improve the digital competence of future health care professionals.

PURPOSE

The aim of this study was to describe the attitudes of medical and nursing students towards digital health based on self-evaluation as well as to compare the differences in perceptions between the two student groups.

METHODS

A cross-sectional study was conducted as an online survey using the Webropol in April 2021 at the University of Oulu and Oulu University of Applied Sciences in Finland. The survey questionnaire consisted of seven background questions and 16 statements on a five-point Likert scale (fully disagree to fully agree) to survey student attitudes towards eHealth, and their digital capabilities.

RESULTS

A total of 250 medical and nursing students were invited to participate in the study and 170 of them took the survey (response rate 68 %). Of those answered, 38 % (n = 64) were nursing and 32 % (n = 106) medical students. Students generally had a positive attitude towards eHealth and health care digitalisation. The differences in perceptions and preparedness between medical and nursing students were surprisingly small in the two student groups. There was a statistically significant difference between the two groups in three out of 16 statements: these were related to changes in the roles of health care professionals and patients as well as the students' knowledge of information contained in the national patient portal.

CONCLUSIONS

The results of this study provide a good starting point for further harmonisation of the curriculum for both health professional groups regarding the teaching of eHealth and telemedicine.